GEFCom 2017 was a global energy forecasting competition in which 177 academic and commercial teams from around the world competed to accurately predict the electricity load for over 100 different clients for the coming year. Tangent Works won first place out of 177 teams in the qualifying round and 2nd place among 12 finalists in the final round - all based on models generated automatically by TIM.
ANDRITZ is a global supplier of plants, equipment, and services for hydropower stations, pulp and paper, metalworking, steel and other industries. In 2017, they held a hackathon event for machine learning startups to apply their latest technology to a series of challenges in predictive maintenance and anomaly detection. The Tangent Works team won 4 out of 5 challenges and took 1st place in the overall competition among 7 competing companies.
The industrial process AI specialists of Wizata joined forces with the time-series machine learning experts of Tangent Works. Together, we provide a dynamic and scalable AI solution for advanced time series forecasting and anomaly detection that is fully integrated in the production process to automatically predict temperature deviations on the Edge.
Learn MorePrimex selected Tangent Works as the preferred machine learning solution for power plant performance optimization across the United States.
Learn MoreStekker.app manages the charging for a fleet of electric vehicles, to save on electricity costs and CO₂ emissions. The company uses TIM’s InstantML capabilities to predict the most economical times to recharge e-vehicles using solar and wind energy.
Learn MoreCompanies worldwide are struggling to forecast demand due to Covid-19. Enterprise Rent-A-Car started looking for a solution that can help with its supply and demand challenges of their car rental fleet in real-time.
Learn MoreEGSSIS NV (part of the Energy One Group) is a European service provider specializing in shipping and trading services and related software to help companies manage and optimize their gas and power energy portfolios. They chose TIM to provide individualized energy forecasts to their clients to help them make informed buying decisions and avoid imbalance costs. TIM was integrated with EGSSIS’s internally-developed software platform to make forecasting a seamless part of the system.
Learn MorePowerhouse offers a self-service energy management portal that enables customers to purchase energy at discount prices and accelerate the shift to renewable sources. Accurate forecasting is a key requirement for both energy buyers and sellers but few of Powerhouse’s customers had the ability to generate forecasts internally. After an extensive evaluation Powerhouse selected TIM to be an integral part of their portal so that customers could simply upload their historical data and TIM would provide a forecast in seconds.
SEPS the Slovak transmission grid operator, integrates the country’s transmission network with Europe’s leading electricity transmission networks. SEPS has implemented TIM for day-ahead and intraday forecasting of active power losses, so called “technical losses”. This allows a more accurate prediction of electricity required to cover transmission losses. Better predictions result and better timed market nominations which ultimately leads to reducing costs. TIM has been seamlessly integrated into the system landscape of SEPS as a module of the Energy Management Execution System and requires little maintenance or involvement of machine learning experts.
Elia is the Belgian based transmission grid operator. Elia used TIM to benchmark its current models for system imbalance forecasting and end of month client settlement forecasting. In both cases, TIM generated in models in seconds rather than the days and weeks of costly engineering times required to build the models by traditional methods in the past. The models generated were generated automatic and as a human readable formula allowing Elia engineers to interpret the dynamics hidden in the data. TIM offers the potential to save huge on engineering times and opens the opportunity to use predictive models in area’s not considered in the past due to the high cost of engineering.